1. Field of the Invention
The present invention relates generally to linear frequency chirped lasers, and more particularly to a linear frequency chirped laser that is either helical phase plate driven or spiral phase wheel driven.
2. Description of the Prior Art
The most common system used in the prior art to cause a laser to rapidly "chirp" uses an electro-optic crystal within the laser resonator. An increasing voltage applied to the crystal causes its index of refraction to change, changing the optical length of the resonator cavity and causing the resonator's longitudinal modes to tune. The specifics of such a prior art system will be discussed for waveguide CO.sub.2 lasers.
The disadvantages and problems of the prior art include, but are not limited to, the following. The electro-optic crystals are typically cadmium telluride (CdTe) of 2 mm.times.2 mm in cross-section and 5 cm long. The crystals must be polished on all surfaces, with the 2 mm.times.2 mm ends being of optical quality and including anti-reflection coatings. The electro-optic crystals of cadmium telluride are very expensive and very fragile.
The electro-optic effect requires a ramp of 0 to 1600 volts across the 2 mm crystal width. The ramp must be very accurately linear, and it must be swept in, typically, 3 to 30 .mu.sec. This is a very difficult requirement combining both high voltages and radio frequencies (RF). Furthermore, the crystal has an index of refraction of 2.7, so its optical length is 13.5 cm. This length, plus the length for the laser gain, requires that the free spectral range is limited to 500 MHz, and the laser cannot reach the chirp amplitude of 1 GHz desired for some applications without adding longitudinal mode suppression to an already very complex device. The present invention adds less than 2 cm to the optical length of the laser, so the laser can easily reach the 1 GHz chirp amplitude.
The electrode connections must be made between two of the 2 mm.times.5 cm faces of the crystal, and the high voltage RF power must not arc to other parts of the crystal holder or flash-over from one electrode to the other.
Residual strain in the crystal or strain induced by the crystal holder causes transverse variation in the index of refraction, which will distort the optical beam and lead to optical loss within the resonator and a degraded beam quality of the laser output. The strain will also cause birefringence, which will convert one polarization of the light into the other. This will cause further optical distortions because the electro-optic effect is different for the different polarizations. It will also increase the optical losses because some of the optical components will be polarization selective (the spectral line-selecting diffraction grating, for example, and, possibly, the waveguide bore).
Electro-optic crystals are also piezo-electric, so the applied voltage ramp also causes a dimensional change. Because the voltage ramp occurs rapidly, a spectrum of acoustic waves are generated within the crystal. The acoustic energy will fracture the crystal if it is not removed. Even if attenuated enough so that fracture of the crystal will not occur, the acoustic energy will still induce dynamic strain-optic effects. Therefore, this energy must be removed to a very high degree of completeness.
The crystal also absorbs some optical power, which shows up as heat that must also be removed. The heat input is not uniform across the crystal, but is proportional to the optical intensity, which is most intense on the optical axis, farthest from the cooling walls. The absorbed power sets up a temperature gradient within the crystal, which causes index of refraction gradients and mechanical strains, which in turn cause strain-optic effects. These effects limit the circulating optical power and the achievable laser power and beam quality.
The electro-optic crystal must be aligned with the waveguide bore to within very exacting tolerances in both offset and angle, or the insertion losses of the modulator become very large and the optical quality of the laser output drops. This requires a precision adjustable crystal holder and a difficult alignment procedure. The crystal holder must also absorb the acoustic energy, route the high voltage RF, and provide the cooling, all without straining the crystal. The very best electro-optic modulators should have insertion losses of about 6% round-trip, and more typically they are about 10%. The present invention has an insertion loss of less than 1% round-trip.
The chirp linearity and repeatability limit the resolution of a radar system using a chirped laser transmitter. The linearity of the electro-optic driven chirp depends mainly on the linearity of the voltage ramp with time and the piezo-electric/strain optic dynamic disturbances in the index of refraction. Other effects such as the thermal and strain-induced index of refraction gradients and birefringence will also contribute to non-linearities in the resulting optical frequency chirp. Similarly, the repeatability of the chirp slope depends not only on the repeatability of the voltage drive, but also on the dynamic strain-optic effects and the temperature dependence of the electro-optic coefficient. The prior art systems can keep deviations from linearity of the chirp ramp to 1% of the chirp amplitude, and the value of the chirp slope can be maintained constant to approximately 1% from chirp-to-chirp. The present invention very significantly improves these numbers by at least a factor of 100! The system of the present invention solves substantially all of the problems of the prior art while avoiding its shortcomings.